JP2002328885A - Clustering system and method, data processor, program for clustering and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the processing loads of a server, to shorten time required for original service provision, to lower the costs of the entire system and to improve the reliability of network communication. SOLUTION: An authentication request part 23 requesting user authentication from a client terminal 1 to one of a plurality of servers and an authentication re-request part 24 switching to the other server and requesting the provision of a service again when a response to the request is not returned are provided. The presence/absence of fault generation in the server is judged on a client side and the server is switched by processing on the client side in the case that a fault is generated. Thus, the need of performing the processings of heartbeat communication and the monitoring of an opposite server, etc., for clustering in the server is completely eliminated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a clustering system and method, a data processing apparatus, a clustering program, and a recording medium, and more particularly, to a logically independent plural server capable of providing the same service to a client. It is suitable for use in a clustering system that is made to function as the above one server.

[0002]

2. Description of the Related Art In a client-server system in which a server and a client using a service provided by the server are connected via a network, when a server provides a service to a client,
When a server fails, it is important to ensure the redundancy of the server against the failure so that the client can continue to provide services without making the client aware of the failure.

For this purpose, a plurality of servers capable of providing the same service are prepared, and when a normally used master server cannot provide a service due to a failure, the server is switched to another slave server to provide the same service. There is provided a clustering system for performing this.

FIG. 7 is a diagram showing a configuration of a conventional clustering system. In FIG. 7, the first server 51
, A second server 52 and a client 53 are connected via a network 54. First server 51
And the second server 52 provide the same service to the client 53, but normally the first server 51 operates as a master server, and the second server 52 is in a standby state as a slave server.

[0005] Different IP addresses are assigned to the first server 51 and the second server 52, respectively. Therefore, the client 53 is the first server 5
When a failure has not occurred in the first server 51, the network 54 is accessed with the IP address of the first server 51 as a destination, and the service is provided from the first server 51.

The operation system (OS) 56 of the first server 51 and the OS 57 of the second server 52
The mutual operation status is constantly monitored by heartbeat communication established by a communication path 55 such as a LAN (Local Area Network). Therefore, the second server 52
In the standby state, only the hardware and the OS 57 required for heartbeat communication and monitoring of the first server 51 are in an active state.

In such a configuration, the first server 5
When the first server 51 fails to provide a service to the client 53 due to a failure, the second server 52 recognizes this state by heartbeat communication, takes over the IP address of the first server 51, and takes over its own IP address. Set as Thereby, the failed first server 51
Instead, the second server 52 starts operating.

By this switching, the client 53 sends the same IP address as the destination to the second server 5
2 can be accessed and the service can be provided from the second server 52. That is, the client 53
On the side, the service can be provided as before without any awareness that a failure has occurred in the first server 51.

[0009]

However, in the above-mentioned conventional clustering system, the servers 51, 52
Must always perform heartbeat communication for clustering and monitoring of the partner server in addition to the original processing for providing a service to the client 53.
Therefore, there is a problem that the processing load on the servers 51 and 52 becomes extremely large.

[0010] In addition, since the limited processing capacity of the servers 51 and 52 is consumed by the clustering, there is a problem that the time required for providing the original service becomes long.

Further, in the conventional clustering system, it is necessary to provide a mechanism for heartbeat communication and monitoring of the partner server in both servers 51 and 52.
Not only does the system become complicated, but also the cost of the entire system increases.

Further, even if a failure has not occurred in the first server 51 due to a malfunction of the heartbeat communication, the second server 51 is not required.
Server 52 may be switched to the operating state, and an IP address collision may occur. In this case, a plurality of servers 51 and 52 correspond to one IP address, and there is also a problem that the reliability of network communication is reduced such that communication of data cannot be performed due to an unknown destination of data.

SUMMARY OF THE INVENTION The present invention has been made to solve such a problem, and it is an object of the present invention to reduce the processing load on a server, to shorten the time required for providing an original service, It is intended to realize the entire system at low cost and to improve the reliability of network communication.

[0014]

According to the clustering system of the present invention, a plurality of physically independent servers and a client which receives the same service provided by the plurality of servers from any one of the servers are connected on a network. ,
In a clustering system in which the plurality of servers logically function as one server, the client is a requesting unit that requests any one of the plurality of servers to provide the service; Response detecting means for detecting whether or not a response has been returned from any one of the plurality of servers in response to the request for providing the service, and detecting that the response has not been returned by the response detecting means. Re-requesting means for re-requesting the provision of the service to another server when the request is made.

[0015] In another aspect of the present invention, the response detecting means does not receive the response if the response is not returned within a predetermined time after issuing the request for providing the service by the requesting means. It is characterized by judging that.

In another aspect of the present invention, a plurality of physically independent servers and a client which receives the same service provided by the plurality of servers from any one of the servers are connected on a network, and the plurality of servers In a clustering system that logically functions as one server, the client includes a request unit for simultaneously requesting the plurality of servers to provide the service, and a request for providing the service by the request unit. Correspondingly, of the plurality of responses returned from the plurality of servers, only the earliest response is adopted, and thereafter, the server which has returned the earliest response is selectively provided with the service. And a server selecting means for performing the operation.

According to another aspect of the present invention, there is provided a notifying means for notifying an error when the response is not returned at all within a predetermined time after the request for providing the service is issued by the requesting means. Features.

In another aspect of the present invention, the server is an authentication server that performs a process related to user authentication for permitting only a specific user to access the network or a system on the network. And

Further, according to the clustering method of the present invention, a plurality of physically independent servers and a client receiving the same service provided by the plurality of servers from any one of the servers are connected on a network, and the plurality of servers are connected to each other. Requesting the client to request one of the plurality of servers to provide the service, wherein the client requests the service to be provided to any one of the plurality of servers. A response detecting step of detecting whether a response is returned from any one of the plurality of servers; and, when detecting that the response is not returned, detecting a response of the service to another server. Requesting the provision again.

In another aspect of the present invention, a plurality of physically independent servers and a client that receives the same service provided by the plurality of servers from any one of the servers are connected on a network, and the plurality of servers Requesting the client to request the plurality of servers to provide the service at the same time, and the plurality of servers corresponding to the request to provide the service. A server selecting step of adopting only the earliest response among a plurality of responses returned from the server, and thereafter selectively receiving the service from the server that has returned the earliest response; and It is characterized by having.

Further, the data processing apparatus of the present invention is characterized in that a plurality of physically independent servers and a client receiving the same service provided by the plurality of servers from any one of the servers are connected on a network, A data processing device used as the client in the clustering system in which the server logically functions as a single server, wherein the data processing device requests one of the plurality of servers to provide the service. Means, response detecting means for detecting whether a response has been returned from any one of the plurality of servers to the request for providing the service by the requesting means, and returning the response by the response detecting means. Requesting another server to provide the above service again when it is detected that Characterized by comprising and.

In another aspect of the present invention, a plurality of physically independent servers and a client that receives the same service provided by the plurality of servers from any of the servers are connected on a network, and the plurality of servers are connected to each other. Requesting means for simultaneously requesting the plurality of servers to provide the service, in the clustering system in which the server functions as one logical server. Of the plurality of responses returned from the plurality of servers in response to the request for provision of the service described above, only the earliest response is adopted, and thereafter, selectively from the server that has returned the earliest response. A server selecting means for receiving the provision of the service.

Further, the clustering program of the present invention requests a client connected via a network to provide any one of a plurality of physically independent servers providing the same service. Requesting means, a response detecting means for detecting whether a response has been returned from any one of the plurality of servers to a request for provision of the service by the requesting means, and the response detecting means This is for causing the computer to function as re-requesting means for re-requesting the provision of the service to another server when it is detected that the service is not returned.

Another aspect of the present invention is a requesting means for simultaneously requesting a plurality of physically independent servers providing the same service from a client connected via a network to provide the service, and Of the plurality of responses returned from the plurality of servers in response to the request for providing the service by the requesting means, only the earliest response is adopted, and thereafter, the server that has returned the earliest response is selected. This is for causing the computer to function as a server selecting means for receiving the provision of the service.

Further, a computer-readable recording medium according to the present invention is characterized in that a program for causing a computer to function as each means described in claim 9 or 11 is recorded.

[0026]

(First Embodiment) A first embodiment of the present invention will be described below with reference to the drawings. FIG.
FIG. 1 is a diagram illustrating a configuration of an entire network system to which a clustering system according to an embodiment is applied.

In FIG. 1, reference numerals 1, 2, and 3 denote client terminals composed of personal computers and the like, 9 denotes a file server, 10 denotes a mail server, 11 denotes a personnel / payroll server, and 12 denotes an accounting / financial server. 20 so that they can communicate with each other.

The file server 9 performs processes such as file transfer, deletion, and directory operation. The mail server 10 transmits an e-mail based on a request from the terminals 1, 2, and 3, stores the received e-mail, and
It performs processing such as handing over when there is an inquiry from 2 and 3. The personnel and salary server 11 performs various processes related to personnel and salary in a company. The accounting and finance server 12
Performs various accounting and financial processes within a company. In addition,
Known servers can be used for these various servers 9 to 12, and a detailed description of the processing content is omitted here.

Reference numeral 8 denotes a router, which is installed at an appropriate position on the network 20. Data transmitted from a certain computer over the network 20 must be
Delivered to the target computer via. The router 8 refers to the routing information (routing table) of the router 8 based on the destination IP address in the IP header, determines the next node to be transferred, and transfers the data.

13a is a first authentication server, and 13b is a second authentication server.
Authentication server, which is installed between the first to third terminals 1 to 3 and the personnel / payment server 11 and the accounting / finance server 12. These authentication servers 13a, 13b
User authentication based on authentication information sent from the first and second terminals 1 and 2 in order to permit only specific users to access the personnel / payment server 11 and the accounting / finance server 12 Related processing.

The first authentication server 13a and the second authentication server 13b provide the same authentication service to the first and second terminals 1 and 2, but usually the first authentication server 13a does not It operates as a master server, and the second authentication server 13b is on standby as a slave server.
If the first authentication server 13a fails for some reason, the second authentication server 13b starts operating instead.

Different IP addresses are assigned to the terminals 1 to 3 and the servers 8 to 12, 13a and 13b on the network 20. Of course, the first
The IP address of the authentication server 13a is different from the IP address of the second authentication server 13b. Therefore, when no failure has occurred, the first and second terminals 1 and 2 access the network 20 with the IP address of the first authentication server 13a as a destination, and the first authentication server 13a sends a service request. Receive the offer.

The first terminal 1 is connected to a dedicated reader 4 of an IC card 5. The IC card 5 includes, for example, personnel
Authentication information (user status information, biometric information such as a fingerprint, voice, face, and the like) relating to a user who has access to the pay server 11 is stored. The user of the first terminal 1 can freely access the file server 9 and the mail server 10 and can also access the personnel / payroll server 11 by receiving user authentication using the IC card 5. become.

When the user of the first terminal 1 accesses the personnel / payroll server 11, the user first inserts the IC card 5 into the dedicated reader 4 and causes the first terminal 1 to read his / her authentication information. The first terminal 1 sends the read authentication information to the first authentication server 13a via the router 8. If no failure has occurred, the first authentication server 13a
The authentication information sent from the terminal 1 is checked, and if it is correct, the access to the personnel / payroll server 11 is permitted.

The second terminal 2 is connected to a dedicated reader 6 of an IC card 7. The IC card 7 stores, for example, authentication information (user status information, biometric information such as fingerprints, voices, faces, etc.) relating to users who have access to the accounting / finance server 12. The user of the second terminal 2 has a file server 9
And the mail server 10 can be freely accessed, and the user can be authenticated using the IC card 7 to access the accounting / finance server 12.

When the user of the second terminal 2 accesses the accounting / financial server 12, first, the user inserts the IC card 7 into the dedicated reader 6 and causes the second terminal 2 to read his / her authentication information. The second terminal 2 sends the read authentication information to the first authentication server 13a via the router 8. If no failure has occurred, the first authentication server 13a
The authentication information sent from the terminal 2 is checked, and if it is correct, the access to the accounting / finance server 12 is permitted.

The third terminal 3 does not have a function for receiving user authentication. That is, the user of the third terminal 3 includes the personnel / payroll server 11 and the accounting / financial server 1
2 has no access right to the file server 9 and the mail server 10.

Here, the first and second terminals 1, 1
Although the dedicated readers 4 and 6 for the IC cards 5 and 7 are provided externally, the first and second terminals 1 and 2 themselves may have the function of reading the IC cards 5 and 7.
Although the IC cards 5 and 7 are used for user authentication here, the present invention does not particularly limit the user authentication method. For example, another user authentication technique such as a password may be used.

Here, those requiring user authentication for access are the personnel / payment server 11 and the accounting / financial server 12, but are not limited to these servers. For example, user authentication may be used as a prerequisite for access to another server (not shown), the file server 9 or the mail server 10, or a host computer (not shown).

FIG. 2 is a block diagram showing an example of a functional configuration of the first terminal 1. Note that the second terminal 2 is also the first terminal 1
Since the configuration is the same as that described above, illustration is omitted here. In FIG. 2, reference numeral 21 denotes a communication unit which performs processing related to transmission and reception of data via the network 20. Reference numeral 22 denotes an authentication processing unit, and the first and second authentication servers 13a, 13a
In cooperation with 3b, processing related to user authentication is performed.

The authentication processing section 22 takes in the authentication information in the IC card 5 read by the dedicated reader 4 and sends it to the first and second authentication servers 13a, 13a via the communication section 21.
3b. In addition, it also has a function of capturing and holding authentication permission information sent from the first and second authentication servers 13a and 13b via the communication unit 21. It is possible to access the personnel / payroll server 11 only while the authentication processing unit 22 holds the authentication permission information.

As shown in FIG. 2, the authentication processing section 22 has an authentication request section 23 and an authentication re-request section 24 as its functional configuration. The authentication request unit 23 requests the user authentication by transmitting the authentication information in the IC card 5 read by the dedicated reader 4 to the first authentication server 13a, which is a master server, by the communication unit 21.

Further, the authentication re-request unit 24 is provided with the authentication request unit 2
The first authentication server 13 responds to the authentication request issued by
If no response is returned from a, the first authentication server 13a is determined to have failed and the IC card 5 read by the dedicated reader 4 to the second authentication server 13b, which is a slave server, is read. Is transmitted by the communication unit 21 to request the user authentication again.

Reference numeral 25 denotes a response detecting unit, which is an authentication requesting unit 23.
Detects whether a response to the user authentication request made by the first authentication server 13a is returned from the first authentication server 13a to the communication unit 21. For example, if no response is returned within a certain period of time after the authentication requesting unit 23 issues a request for user authentication to the first authentication server 13a, it is determined that there is no response, and that fact is determined. Tell 24.

FIG. 3 is a diagram showing a general communication protocol. When sending data from the sender to the receiver,
First, a signal START is transmitted from the transmission side to the reception side to start data transmission. In response to this, when the response signal Ack is returned from the receiving side to the transmitting side, data Data is transmitted from the transmitting side to the receiving side. When the data transmission is completed, a signal END indicating that the data transmission has been completed is transmitted from the transmission side to the reception side. In response to this, a response signal Ack is returned from the receiving side to the transmitting side, thereby completing a series of data communication.

In such a communication protocol, when the transmitting side is the first terminal 1 and the receiving side is the first authentication server 13a, first, the authentication request unit 23 of the first terminal 1
Sends a signal START requesting user authentication to the first authentication server 13a with the IP address as a destination.

If the first authentication server 13a has not failed and is operating normally, the first authentication server 13a
3a returns a response signal Ack to the first terminal 1 with its own IP address attached to the user authentication request. When the response detection unit 25 detects the reception of the response signal Ack,
Thereafter, the first terminal 1 establishes communication with the first authentication server 13a, sends authentication information to the first authentication server 13a, and receives user authentication service.

On the other hand, when a failure has occurred in the first authentication server 13a, the first authentication server 13a
Does not return a response signal Ack to the user authentication request sent from the terminal 1 of the. Therefore, in this case, the response detection unit 25 detects that the response signal Ack has not been returned within a certain period of time after issuing the user authentication request, and accordingly, the authentication re-request unit 24 sets the second authentication server 13b to the IP address to send a signal START in order to re-request user authentication.

At this time, if the second authentication server 13b has not failed and is operating normally, the second authentication server 13b responds to this user authentication request with its own
A response signal Ack is returned to the first terminal 1 with the P address. When the response detection unit 25 detects the reception of the response signal Ack, the first terminal 1 thereafter transmits the second authentication server 13
b), and sends authentication information to the second authentication server 13b to receive a user authentication service.

FIG. 4 is a flowchart showing the operation of the first terminal 1 in this case. In FIG. 4, first, the authentication request unit 23 sends a signal START requesting user authentication to the first authentication server 13a with the IP address as a destination (step S1). Then, the response detection unit 25 determines whether or not the response signal Ack is returned from the first authentication server 13a within a predetermined time after transmitting the signal START (step S2).

The response signal Ac from the first authentication server 13a
If k is returned, the authentication processing unit 22 returns to the first
Establishes communication with the authentication server 13a (step S
3) The authentication information is sent to the first authentication server 13a to receive the user authentication service (step S6).

On the other hand, if a failure has occurred in the first authentication server 13a and the response signal Ack has not been returned from the first authentication server 13a within a certain period of time, the authentication re-request section 24 sets the START requesting user authentication to the second authentication server 13b with the IP address as a destination
Is sent (step S4).

In response, the second authentication server 13b
Returns a response signal Ack to the first terminal 1, the authentication processing unit 22 establishes communication with the second authentication server 13b (step S5), and the second authentication server 13
b, the authentication information is sent to the user, and the user authentication service is provided (step S6).

When both the first authentication server 13a and the second authentication server 13b are out of order, the second authentication server 13b sends the first terminal 1 to the user authentication request in step S4. No response signal Ack is returned. To cope with this case, step S4
It may be determined whether a response signal Ack has been received between step S5 and step S5, and if not, an error notification may be made.

As described above in detail, in the first embodiment, the client determines whether a failure has occurred in the server, and if a failure occurs, the client changes the destination IP address of the authentication request. Server switching. Thereby, in the first and second authentication servers 13a and 13b,
There is no need to perform any processing such as heartbeat communication or monitoring of the partner server for clustering.

Therefore, the processing load on the first and second authentication servers 13a and 13b can be remarkably reduced. In addition, the processing capacity previously spent for heartbeat communication for clustering and monitoring of the partner server can be allocated to the original processing of the authentication servers 13a and 13b to concentrate on the user authentication processing.
The hardware resources 3a and 13b can be effectively used. Therefore, even if hardware having the same specifications as in the past is used, the time required for the user authentication process can be reduced.

In recent years, client terminals 1 and 2
The hardware specifications of personal computers are becoming higher and they are comparable to server machines. In the present embodiment, the hardware resources of the client terminals 1 and 2 can be effectively used by causing the client terminals 1 and 2 to perform a part of the clustering process conventionally performed by the server.

As a process for the client terminals 1 and 2 to detect the occurrence of a failure in the first authentication server 13a, when the client terminals 1 and 2 actually request user authentication, the response signal Ack All you have to do is see if you have received it. Therefore, the clustering mechanism works only when user authentication, which is the original process, is performed, and the waste of communication traffic is reduced as compared with the conventional system that constantly connects to the network and monitors heartbeat communication and the other server. Can also.

According to the present embodiment, there is no need to incorporate any mechanism for heartbeat communication or monitoring of the partner server in the first and second authentication servers 13a and 13b. Simply connect two authentication servers to the network 20. Thus, the system can be simplified and the cost of the entire system can be reduced.

Further, since the heartbeat communication does not need to be performed, the first authentication server 13a
The inconvenience that the second authentication server 13b is switched to the operating state even if no failure has occurred can be avoided. Therefore, collision of IP addresses can be prevented, and the reliability of network communication can be improved.

(Second Embodiment) Next, a second embodiment of the present invention will be described with reference to the drawings. The configuration of the entire network system to which the clustering system according to the second embodiment is applied is the same as that of FIG. However, the first
Regarding the second authentication server 13a and the second authentication server 13b, there is no distinction as to which is the master server and which is the slave server.

FIG. 5 is a block diagram showing an example of the functional configuration of the first terminal 1 (the same applies to the second terminal 2) in the second embodiment. In FIG. 5, reference numeral 31 denotes a communication unit, which performs processing related to transmission and reception of data via the network 20. Reference numeral 32 denotes an authentication processing unit that performs processing related to user authentication in cooperation with the first and second authentication servers 13a and 13b.

The authentication processing section 32 fetches the authentication information in the IC card 5 read by the dedicated reader 4 and sends the first and second authentication servers 13a, 13a through the communication section 31.
3b. In addition, it also has a function of capturing and holding authentication permission information sent from the first and second authentication servers 13a and 13b via the communication unit 31. It is possible to access the personnel / payroll server 11 only while the authentication processing unit 32 holds the authentication permission information.

As shown in FIG. 5, the authentication processing section 32 has an authentication request section 33 and a server selection section 34 as its functional configuration. The authentication requesting unit 33 includes the first authentication server 1
It requests user authentication to both 3a and the second authentication server 13b at the same time. Note that the term “simultaneous” here does not have to be strictly simultaneous, and for example, a case where an authentication request is sent to the second authentication server 13b immediately after an authentication request is sent to the first authentication server 13a. included.

The server selection unit 34 sends the first terminal 1 of the plurality of response signals Ack returned from the first and second authentication servers 13a and 13b in response to the user authentication request by the authentication request unit 33. Only the earliest response signal Ack is adopted, and thereafter, communication is selectively established with the server that has returned the earliest response signal Ack, and user authentication is received from the server.

As described above with reference to FIG. 3, the first and second authentication servers 13a and 13b receiving the user authentication request from the first terminal 1 operate normally without any failure. If so, the mobile terminal returns the response signal Ack to the first terminal 1 with its own IP address. Therefore, the server selection unit 34 determines which server should be selected based on the IP address attached to the response signal Ack returned first from the authentication servers 13a and 13b to the communication unit 31. It is possible.

Reference numeral 35 denotes a response detection unit, and the authentication request unit 33
Detects whether a response signal Ack is returned from the authentication servers 13a and 13b to the communication unit 31 in response to the user authentication request made by the first and second authentication servers 13a and 13b. Which authentication server 13
If no response signal Ack is returned from a and 13b, there is a possibility that a failure has occurred in both authentication servers 13a and 13b, and an error notification to that effect is given. On the other hand, when the response signal Ack is returned, the server selection unit 34 selects a server.

FIG. 6 is a flowchart showing the operation of the first terminal 1 in this case. In FIG. 6, first, the authentication requesting unit 33 converts the first and second authentication servers 13 a and 13
b simultaneously sends a signal START requesting user authentication to each IP address (step S1).
1). Then, the response detection unit 35 determines whether or not the response signal Ack is returned from the first and second authentication servers 13a and 13b within a predetermined time after transmitting the signal START (step S12).

When the response signal Ack is returned, the server selecting unit 34 adopts only the response signal Ack that has arrived first, and thereafter communicates with the authentication server that has returned the earliest response signal Ack. Establish communication selectively (step S1
3) The authentication information is sent to the authentication server to receive the user authentication service (step S14). On the other hand, when the response signal Ack is not returned from any of the authentication servers 13a and 13b, both the authentication servers 13a and 13b
An error notification is given to the effect that a failure may have occurred in 13b (step S15).

As described in detail above, in the second embodiment, the client issues an authentication request to all the authentication servers, and performs user authentication with the authentication server which normally returns a response signal without any trouble. Is executed. At this time, when response signals are returned from a plurality of authentication servers, the server that transmitted the response signal that has reached the client first is selected to perform user authentication. As a result, the first and second authentication servers 13a and 13b do not need to perform any processing such as heartbeat communication or monitoring of the partner server for clustering.

Therefore, the processing load on the first and second authentication servers 13a and 13b can be remarkably reduced. In addition, the processing capacity previously used for heartbeat communication for clustering and monitoring of the partner server can be allocated to the original processing of the authentication servers 13a and 13b to concentrate on the user authentication processing.
The hardware resources 3a and 13b can be effectively used. Therefore, even if hardware having the same specifications as in the past is used, the time required for the user authentication process can be reduced.

Further, a part of the clustering processing conventionally performed by the server is performed by the client terminals 1 and 2 having hardware specifications comparable to those of the server machine, so that the hardware resources of the client terminals 1 and 2 are reduced. It can be used effectively.

Further, since the clustering mechanism can be operated only when the user authentication, which is the original processing, is performed, the communication can be performed in comparison with the conventional method in which the network is constantly connected to monitor the heartbeat communication and the partner server. It can also reduce traffic waste.

Further, according to the present embodiment, there is no need to incorporate any mechanism for heartbeat communication or monitoring of the partner server in the first and second authentication servers 13a and 13b. Simply connect two authentication servers to the network 20. Thus, the system can be simplified and the cost of the entire system can be reduced.

Since heartbeat communication does not need to be performed, collision of IP addresses due to malfunctions can be prevented, and reliability of network communication can be improved.

Further, according to the present embodiment, it is not necessary to determine in advance a master server to be used at normal times and a slave server to be used when a failure occurs in the master server, so that the processing for clustering is simplified. be able to. Also, the network 20 and each authentication server 13a,
13b, the response signal Ack first.
May change each time, and it is not necessary to abuse one of the authentication servers as a master server, so that the durability of the authentication servers 13a and 13b can be improved.

Further, since the authentication processing is performed with the authentication server which has returned the response signal Ack first, the time from when the request is issued to when the authentication processing is actually performed is not lengthened. This can reduce user's irritability.

Further, for example, when the first authentication server 13a performs user authentication relating to access to the personnel / payment server 11 from the first terminal 1, the second terminal 2 When a user authentication request for access is made, the second authentication server 13b that has not performed the user authentication process can perform the user authentication by returning the response signal Ack first. That is, the first and second authentication servers 13a, 13a
b can execute the user authentication process at the same time, and the two authentication servers 13a and 13b can be effectively used.

The clustering systems according to the first and second embodiments described above are mainly composed of a CPU of a client computer, an MPU, a RAM, a ROM, or the like. It can be realized by operating. Therefore, the present invention can be realized by recording a program that causes a computer to perform the above function on a recording medium such as a CD-ROM and reading the program into the computer.

As a recording medium for recording the above program, a floppy (registered trademark) disk, a hard disk, a magnetic tape, an optical disk, a magneto-optical disk, a DVD, a nonvolatile memory card, and the like can be used in addition to the CD-ROM. In addition, the above program is transferred to network 2
0 may be downloaded from another computer.

Further, not only the functions of the above-described embodiments are realized by the computer executing the supplied program, but also the OS (Operating System) or other application software running on the computer. When the functions of the above-described embodiment are realized in cooperation with the computer, or when all or a part of the processing of the supplied program is performed by a function expansion board or a function expansion unit of a computer, the functions of the above-described embodiment are realized. Such a program is also included in the embodiment of the present invention.

Further, in order to use the present invention in a network environment, all or some of the programs may be executed by another computer.

In the above embodiments, the number of authentication servers is two, but a similar process may be performed using a larger number of authentication servers.

Further, in the above embodiment, the authentication server is described as an example of the server that performs the clustering, but the present invention is not limited to this. For example, the same clustering as in the above-described embodiment may be applied to the file server 9, the mail server 10, the personnel / payroll server 11, the accounting / finance server 12, and the like connected on the network 20 in FIG. .

However, in the case where clustering is performed for a server whose contents of the accompanying database are updated as needed, it is necessary to provide the same service as before when a failure occurs and switching is performed. It is necessary to exchange the contents of the database between the servers one by one so that the contents are always the same. Therefore, the present invention is particularly suitable for application to a server that does not require such database replication processing.

In addition, each of the above-described embodiments is merely an example of the embodiment for carrying out the present invention, and the technical scope of the present invention should not be interpreted in a limited manner. It is. That is, the present invention does not depart from the spirit or the main features thereof,
It can be implemented in various forms.

[0087]

As described above, according to the present invention, when a client requests the provision of a service to one of a plurality of servers and a response to the request is not returned, the client requests the other server. In this case, the client is requested to provide the service again, so that it is possible to determine whether or not a failure has occurred in the server on the client side, and if a failure has occurred, the server can be switched by processing on the client side. This eliminates the need for the server to perform any processing such as heartbeat communication or monitoring of the partner server for clustering. Therefore, it is possible to reduce the processing load on the server, shorten the time required to provide the service, reduce the cost of the entire system, improve the reliability of network communication, and reduce waste of communication traffic.

According to another feature of the present invention, a client requests a plurality of servers to provide a service at the same time, and selects a server to which a response to the request arrives first to receive the service. Therefore, the following effects can be obtained in addition to the effects described above.
That is, since it is not necessary to determine in advance a master server to be used during normal operation and a slave server to be used when a failure occurs in the master server, it is possible to simplify the processing for clustering. Further, since it is not necessary to abuse only one of the servers as the master server, the durability of the server can be improved. In addition, since the service is provided from the server that has returned the response first, the time from when the request is issued to when the service is actually provided can be prevented from becoming long. Further, it is possible to provide a service simultaneously with a plurality of servers prepared for clustering, and it is possible to effectively utilize the plurality of servers.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a configuration of an entire network system to which a clustering system according to first and second embodiments is applied.

FIG. 2 is a block diagram illustrating a functional configuration example of a first terminal according to the first embodiment.

FIG. 3 is a diagram showing a general communication protocol.

FIG. 4 is a flowchart showing an operation of the first terminal according to the first embodiment.

FIG. 5 is a block diagram illustrating a functional configuration example of a first terminal according to a second embodiment.

FIG. 6 is a flowchart illustrating an operation of a first terminal according to the second embodiment.

FIG. 7 is a diagram showing a configuration of a conventional clustering system.

[Explanation of symbols]

1,2,3 Client terminal (personal computer) 4,6 Dedicated reader 5,7 IC card 8 Router 9 File server 10 Mail server 11 HR / payment server 12 Accounting / financial server 13a First authentication server 13b Second authentication Server 21 Communication unit 22 Authentication processing unit 23 Authentication request unit 24 Authentication re-request unit 25 Response detection unit 31 Communication unit 32 Authentication processing unit 33 Authentication request unit 34 Server selection unit 35 Response detection unit

Claims (12)

[Claims] 1. A plurality of physically independent servers and a client that receives the same service provided by the plurality of servers from any server are connected on a network, and the plurality of servers are logically connected to one server. In the clustering system configured to function as: a client, a request unit for requesting any one of the plurality of servers to provide the service; and a request unit for providing the service by the request unit. Response detecting means for detecting whether a response has been returned from any one of the plurality of servers; and when detecting that the response has not been returned by the response detecting means, A clustering system comprising: a re-requesting unit for requesting the provision of the service again. 2. The response detecting means determines that there is no response if the response is not returned within a predetermined time after the request for the service is issued by the requesting means. The clustering system according to claim 1, wherein: 3. A plurality of physically independent servers and a client receiving the same service provided by the plurality of servers from any server are connected on a network, and the plurality of servers are logically connected to one server. In the clustering system configured to function as: a client, a request unit that requests the plurality of servers to provide the service at the same time; and the plurality of servers in response to a request to provide the service by the request unit. Server selection means for adopting only the earliest response among a plurality of responses returned from the server and thereafter selectively receiving the service from the server which has returned the earliest response. A clustering system, comprising: 4. A notifying means for notifying an error when the response is not returned at all within a predetermined time after the request for providing the service is issued by the requesting means. The clustering system according to 1. 5. The server according to claim 1, wherein the server is an authentication server that performs a process related to user authentication for permitting only a specific user to access the network or a system on the network. 5. The clustering system according to any one of 4. 6. A plurality of physically independent servers and a client that receives the same service provided by the plurality of servers from any server are connected on a network, and the plurality of servers are logically one server. In the clustering system that is made to function, the client requests the one of the plurality of servers to provide the service, and the client requests the one of the plurality of servers to provide the service. A response detecting step of detecting whether a response has been returned from any one of the servers; and a re-request for requesting another server to provide the service again when it is detected that the response has not been returned. And a clustering method. 7. A plurality of physically independent servers and a client that receives the same service provided by the plurality of servers from any server are connected on a network, and the plurality of servers are logically connected to one server. Requesting the client to simultaneously request the plurality of servers to provide the service; and returning the plurality of servers in response to the request to provide the service. A server selecting step of adopting only the earliest response among the plurality of responses and thereafter selectively receiving the service from the server that has returned the earliest response. Clustering method to use. 8. A plurality of physically independent servers and a client that receives the same service provided by the plurality of servers from any one of the servers are connected on a network, and the plurality of servers are logically connected to one server. A clustering system configured to function as a client; a data processing device used as the client; requesting means for requesting any one of the plurality of servers to provide the service; Response detecting means for detecting whether or not a response has been returned from any one of the plurality of servers in response to a service provision request; and detecting that the response has not been returned by the response detecting means. And a re-requesting means for requesting another server to provide the service again. Data processing equipment. 9. A plurality of physically independent servers and a client which receives the same service provided by the plurality of servers from any one of the servers are connected on a network, and the plurality of servers are logically connected to one server. A data processing device used as the client in the clustering system configured to function as: a request unit for simultaneously requesting the plurality of servers to provide the service; and providing the service by the request unit. Of the plurality of responses returned from the plurality of servers in response to the request, only the earliest response is adopted, and thereafter, the service provided selectively from the server that has returned the earliest response is provided. A data processing apparatus comprising: a server selection unit configured to perform the above operation. 10. Request means for requesting one of a plurality of physically independent servers providing the same service from a client connected via a network to provide the service, wherein the request means Response detecting means for detecting whether a response has been returned from any one of the plurality of servers to the request for providing the service, and detecting that the response has not been returned by the response detecting means. Sometimes, a clustering program for causing a computer to function as re-requesting means for requesting another server to provide the service again. 11. A requesting means for simultaneously requesting a plurality of physically independent servers providing the same service from a client connected via a network to provide said service, and requesting said service by said requesting means. Of the multiple responses returned from the multiple servers in response to the provision request, adopt only the response that arrived first,
Thereafter, a clustering program for causing a computer to function as server selection means for selectively receiving the service from the server that has returned the earliest response. 12. A computer-readable recording medium on which a program for causing a computer to function as each of the means according to claim 9 or 11 is recorded.